do-review-elixir
Pass
Audited by Gen Agent Trust Hub on Sep 13, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill comprises markdown instructions and reference guides for reviewing Elixir, Phoenix, and LiveView code. It provides the agent with structured steps to identify changed files, run standard linting tools (mix format, credo, dialyzer), and verify findings against project conventions.
- [INDIRECT_PROMPT_INJECTION]: The skill analyzes Elixir source code, which serves as an ingestion point for untrusted data. However, the skill implements a robust 'Review Verification Protocol' and multiple 'Hard Gates' that mandate evidence-based verification, searching for references, and checking framework guarantees. These measures effectively mitigate risks associated with processing external code content.
- [COMMAND_EXECUTION]: The skill uses standard development tools like
mix,git, andgrepto analyze code and verify project status. These operations are essential for the skill's function and are used within a safe, predefined context.
Audit Metadata